The paper compares the nature of the textual construction of the Programme for International Student Assessment (PISA) science test items and the Greek school science textbooks. This nature is determined by the interplay of the notions of classification (content specialization) and formality (code specialization) modulated by both the linguistic and the visual expressive modes. The results of the relevant analysis showed that the textual materials employed in PISA and in school textbooks are oppositional in nature. Specifically, while the linguistic mode of the PISA items tends to resemble texts falling within the public domain (nonspecialized content and code), the corresponding visual mode tends to resemble texts of the esoteric domain (specialized content and code) thus familiarizing students with the specialized conventions and ways of representing entities falling within the realm of technoscientific knowledge. On the other hand, school science textbooks tend to employ the linguistic and the visual mode in exactly the opposite way. These differences between PISA science items and school science textbooks could contribute to a disorientation of the students in relation to what is expected in each specific context. This disparity could potentially be one of the factors explaining the low level of Greek students' attainment in PISA. (Contains 10 tables and 4 figures.)